Integrating Robot Manufacturer Perspectives into Legible Factory Robot Light Communications
Alexandra Bacula,Jason Mercer,Jaden Berger,Julie Adams,Heather Knight
DOI: https://doi.org/10.1145/3570732
2023-02-15
ACM Transactions on Human-Robot Interaction
Abstract:In a world with increasing numbers of robots operating in everyday human spaces, the employees at this robotics company are pioneers, with intelligent point-to-point path planning and autonomous transport operations in 150+ factory and warehouse locations in North America. At the time of research, this robotics company consisted of 250 employees. Unlike other industry models, their robots are designed to operate with people in mixed human-machine spaces, yet no HRI style evaluations had previously been run with their robots. As early observers of how factory workers and transport robot interact, across varied job roles ranging from technology design to customer relations, this work sought to leverage employee knowledge and experiences to identify opportunities for improving the communication capabilities of the robots, resulting in the addition of several robot state communications to their initial software set leveraging both employee- and social robotics literature- sourced ideas for communicating with lights. To achieve this a social robotics researcher spent a summer onsite at the robotics company, getting to know their software stack and culture. Her research activities included: (1) a company-wide survey relative to the robot’s light, sound, and motion communications was sent out and analyzed, (2) the development of three new light sets (car-like, sweeping, heartbeat) and five overall states (blocked, at goal, turning, idle), and (3) a user study evaluating the developed light sets relative to the current robot default light patterns, all significantly improving the overall legibility of the targeted robot state communications: at goal, blocked, turning, and idle. Our initial findings advance knowledge in which style of light patterns is best for different communication states, showing that eye-catching lights are best for high urgency states, such as blocked, and subtle lights are best for low urgency states, such as idle. Finally, the latest software release for this robot has deployed a subset of these light patterns to all of their currently operating client sites, i.e., anyone who updates their robots to the latest release will benefit from these research results. This deployment sets the ground for future researchers exploring how end-users at different sites have responded to the new, more communicative light patterns.